Automated 3D Whole-Heart Mesh Reconstruction From 2D Cine MR Slices Using Statistical Shape Model

Area of Science:

• Biomedical Imaging
• Computational Anatomy
• Medical Image Analysis

Background:

• Cardiac magnetic resonance (CMR) imaging is crucial for diagnosing cardiovascular diseases.
• Standard CMR protocols provide 2D slices, often insufficient for comprehensive 3D whole-heart reconstruction, especially for atria.
• Accurate 3D models are vital for advanced cardiac applications.

Purpose of the Study:

• To develop and validate a novel automated method for patient-specific 3D whole-heart mesh reconstruction from limited 2D cine CMR slices.
• To improve the accuracy and completeness of 3D cardiac models derived from CMR data.
• To facilitate clinical applications requiring detailed 3D heart geometry.

Main Methods:

• Utilized a statistical shape model (SSM) for automated 3D whole-heart mesh reconstruction.
• Extracted heart contours from 2D cine CMR slices.
• Optimally fitted the SSM to sparse 3D contours and deformed the mesh to minimize contour-to-surface distance.

Main Results:

• Successfully generated high-quality 3D whole-heart meshes from limited 2D CMR data.
• Achieved an average contour-to-surface distance below the image resolution.
• Demonstrated clinical metrics within acceptable ranges in a cohort of 30 UK Biobank subjects.

Conclusions:

• The proposed automated SSM-based approach enables accurate patient-specific 3D whole-heart mesh reconstruction.
• This method addresses limitations of standard CMR protocols for 3D modeling.
• The generated meshes have significant potential for cardiac diagnosis, treatment planning, and biomedical simulations.